AI & Data Processing Notice

Section 01

What this notice covers

This notice supplements the Privacy Policy and Health & Safety Notice. It applies when you ask GymBuddy or another Revenge Arc feature to generate text, plans, classifications, estimates, summaries, speech, or image-informed feedback.

Section 02

Possible AI features

• Cal AI: meal-photo interpretation and estimated foods, portions, calories, and macronutrients.
• Type & Track: natural-language interpretation of foods, quantities, meals, and nutrition entries.
• GymBuddy and related coaching: conversational assistance, workout planning, exercise suggestions, progress summaries, and general coaching.
• Voice logging: speech recognition, transcription, and conversion of spoken details into app entries.
• Image-, audio-, transcript-, and progress-assisted feedback when you actively select or record that media.

Section 03

Information sent for a request

A request may include your prompt plus the minimum relevant profile, goal, workout, nutrition, progress, conversation, image, audio, or transcript context selected by the feature. Authentication tokens and provider credentials should remain on Revenge Arc servers and should not be exposed to the mobile client.

Do not include another person’s private information without permission. Avoid submitting medical records, government identifiers, financial credentials, or other information that is unnecessary for the feature.

Section 04

Providers and controls

Revenge Arc currently plans to make AI requests through its server to an external AI provider. The production provider, model, contract, data-processing terms, retention, training or data-use settings, region, subprocessors, and deletion controls must be confirmed before launch and reflected here.

Where consent is required for sensitive data or media, the feature should ask before processing. You can choose not to use optional AI features and may request deletion of eligible source data and account-linked AI records.

Section 05

Accuracy and human judgment

Cal AI and all other AI-assisted outputs are probabilistic estimates. They are intended to be useful and may often provide reasonable approximations, but they are not 100% accurate and will never be perfect. They may invent facts, misread media or speech, miss context, rely on incomplete or outdated data, or provide unsuitable exercise or nutrition guidance. Review every output carefully and stop using a recommendation that feels unsafe. AI does not replace a trainer, registered dietitian, therapist, clinician, pharmacist, or emergency service.

Do not treat polished wording, citations generated by AI, a confidence label, repeated answers, or a history of correct outputs as proof of accuracy. Verify information when an error could affect allergies, medical diets, medication, injury, pregnancy, eating-disorder recovery, exercise safety, or another important decision.

Revenge Arc should not use AI output as the sole basis for decisions that create legal or similarly significant effects. Moderation and account-enforcement decisions should include contextual review appropriate to the risk.

Section 06

Feature-specific limitations

• Cal AI can be wrong because a photo may hide ingredients, cooking oils, sauces, portion depth, preparation method, brand, recipe, or serving size; lighting, angle, focus, and occlusion can also change the estimate.
• Type & Track can misunderstand food names, units, quantities, substitutions, and whether a description refers to one serving or an entire recipe.
• Voice logging and transcripts can mishear accents, background noise, brand names, numbers, units, and corrections. Review the text and nutrition entry before saving or relying on it.
• GymBuddy can miss injuries, equipment limits, experience level, recovery needs, contraindications, and other context that affects whether an exercise or plan is appropriate.
• Workout plans, form cues, progression suggestions, and recovery guidance may not fit your body, environment, equipment, technique, or medical history.
• Food databases, barcode records, labels, recipes, and third-party nutrition data can be incomplete, stale, duplicated, or incorrect even when AI is not involved.
• Progress analysis may mistake correlation for cause, overlook missing entries, or infer trends that do not reflect your actual health or performance.

Section 07

Images, audio, and transcripts

Media may contain faces, voices, surroundings, health indicators, or other sensitive information. Upload only media you have the right to use. Transcripts and media analysis may be inaccurate; an image-based meal, body, movement, or progress estimate is not a medical measurement and must not be treated as one. Production storage, temporary-file, transcript, and provider-retention behavior must be verified before these features launch.

Section 08

Safety

• Do not use AI features for emergencies, diagnosis, treatment, medication decisions, or crisis intervention.
• Stop exercising and seek appropriate help for severe pain, fainting, chest pain, breathing difficulty, neurological symptoms, or another urgent concern.
• Seek qualified help before material changes if you are pregnant, injured, managing a condition or medication, or have a history of an eating disorder.
• Report unsafe, discriminatory, or clearly inaccurate output through the support contact below.

Section 09

Retention, deletion, and questions

AI-related retention follows the Privacy Policy and Data Retention & Deletion Policy. Provider-side retention and deletion behavior must be contractually and technically confirmed before publication. Questions, objections, and deletion requests may be sent to revengearchelp@gmail.com.